In a supervised learning, the relationship between the available data and the performance (what is learnt) is not well understood. How much data to use, or when to stop the learning process, are the key questions. In the paper, we present an approach for an early assessment of the extracted knowledge (classification models) in the terms of performance (accuracy). The key questions are answered by detecting the point of convergence, i.e. where the classification model’s performance does not improve any more even when adding more data items to the learning set. For the learning process termination criteria we developed a set of equations for detection of the convergence that follow the basic principles of the learning curve. The developed solution was evaluated on real datasets. The results of the experiment prove that the solution is well-designed: the learning process stopping criteria are not subjected to local variance and the convergence is detected where it actually has occurred.
COBISS.SI-ID: 16212246
Two-party authenticated key agreement protocols using pairings have gained much attention in the cryptographic community. Several protocols of this type where proposed in the past of which many were found to be flawed. This resulted in attacks or the inability to conform to security attributes. In this paper, we propose an efficient identity-based authenticated key agreement protocol employing pairings which employs a variant of a signature scheme and conforms to security attributes. Additionally, existing competitive and the proposed protocol are compared regarding efficiency and security. The criteria for efficiency are defined in this paper, whereas the criteria for security are defined by the fulfillment of security attributes from literature.
COBISS.SI-ID: 14779926
Software organizations are always looking for approaches that help improve the quality and productivity of developed software products. Quality software is easy to maintain and reduces the cost of software development. The Software Factories (SF) approach is one of the approaches to provide such benefits. In this paper, the quality and productivity benefits of the SF approach were examined and evaluated with an experiment involving two treatments - the traditional and the SF approach. For the purposes of this experiment, the Goal – Question – Metric (GQM) approach was used. Participants were grouped into thirty-two teams. There were sixteen projects available. The results were evaluated and presented through quality and productivity criteria, which were used for the experimental study. The results showed that the Software Factories approach was significantly better than the traditional approach.
COBISS.SI-ID: 16112662
In order to have the greatest treatment impact the early and accurate diagnose of Alzheimer’s disease (AD) is essential. In this paper we present a method for analyzing EEG signals with machine learning approach in order to diagnose AD. We show how to extract features out of EEG recordings to be used with a machine learning algorithm for the induction of AD classification model. The obtained results are very promising.
COBISS.SI-ID: 16417302
In the article a study to use massive parallel programming on general PCs for artificial neural networks (ANN) is presented. Graphic processor units (GPU) on mass-market graphical cards may greatly outperform the general processors for some type of applications, both in computation power and in memory bandwidth. Graphic processor consists of a large number of processing cores that may execute a large number of tasks in parallel. The execution of artificial neural networks is intrinsically parallel problem. Therefore, parallel computational architectures like on GPUs lead to a great improvement of speed. Until recently, the programmers of ANN can harness this processing power only with special prepared graphical applications. What is new is that the newest GPU architectures allow more general approach to ANN programming without taking into consideration graphical aspects of GPUs. One of a general-purpose parallel computing architecture is CUDA (Compute Unified Device Architecture) developed by GPU manufacturer Nvidia. Different aspects of ANN implementation with CUDA are discussed in the paper.
COBISS.SI-ID: 16236054